Being female or male is a fundamental category in human life and contributes to variation in social roles and responsibilities, communication and relationships, economic status, mental and physical health care, and health outcomes. Although patient gender would be expected to play a similarly important role in experiences for cancer patients at the end of life, surprisingly little research has focused on this issue. Preliminary analyses carried out for this proposal suggest the presence of important gender differences in cancer patients' preferences and physician practices related to communication about the end of life, with women being less likely than men to want prognostic information and, when death is imminent, less likely to be told they are dying. In addition, this preliminary evidence suggests having dependent children is an important moderator of the relationship between gender and preferences for prognostic information, with women with children least likely, and men with children most likely, to want to know their life expectancy. Understanding the specific impact and mechanisms of these gender differences, and how they vary by patient ethnicity, will be important in designing effective interventions and determining best practices of care. The purpose of this study is to use an existing dataset and generate new data to document the presence and impact of gender differences in prognostic communication among an ethnically diverse group of cancer patients at the end of life, and to identify potential mechanisms that underlie these differences. This study draws on data from an ongoing, multisite NCI- and NIMH-funded study of ethnic differences among cancer patients with a prognosis of 6-months or less and their caregivers (n=800 patient-caregiver pairs, P.I. Prigerson),and includes an analysis of existing survey data as well as primary data collection through the addition of new items to surveys of newly-enrolled patients and caregivers (estimated n=100 pairs). The study will address three aims: (1) To identify the presence and impact of gender differences in preferences for physician-patient communication about prognosis; (2) (a) To examine whether and how the effect of gender varies by patient ethnicity, socioeconomic status, and parental status, and (b) To determine whether there are gender differences in coping styles, family support, and psychosocial distress, and whether these account for expected differences in communication preferences; and (3) To explore whether there are differences by patient gender in whether physicians tell patients when death is imminent. The statistical method of structural equations modeling will be used to analyze these data. Findings will contribute to the understanding of clinically important gender differences at the end of life, and provide the evidence base for a long-term program of research in gender differences at the end of life. The National Cancer Institute has identified end-of-life care research as one of its priority areas. Roughly 50% of people with cancer will die of their disease; understanding how to minimize suffering and maximize quality of life at the end of life is a top priority for this population. Although preliminary evidence suggests the presence of important gender differences in psychosocial aspects and communication preferences at the end of life, little is known about the presence, impact, or mechanisms of these differences. Psycho-oncology researchers Kiss and Meryn (2001) have noted that "being male or female is an important basic human variable that affects health and illness throughout life" and "it would be a breakthrough if researchers at major cancer centers designed studies looking at major psychobiological issues in men and women simultaneously." This study will be situated in the Center for Research in Psychosocial Oncology and Palliative Care at the Dana- Farber/Harvard Cancer Center, and will provide a tremendous opportunity for a strong and ongoing interdisciplinary research effort to understand the impact of gender and improve care for men and women with cancer at the end of life. [unreadable] [unreadable] [unreadable]